Corporate Overview

March 2020 NASDAQ: MRNS @MarinusPharma Safe Harbor Statement

To the extent that statements contained in this presentation are not descriptions of historical facts regarding Marinus, they are forward-looking statements reflecting the current beliefs and expectations of management made pursuant to the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. Words such as “may”, “will”, “expect”, “anticipate”, “estimate”, “intend”, “believe”, and similar expressions (as well as other words or expressions referencing future events, conditions or circumstances) are intended to identify forward-looking statements. Examples of forward-looking statements contained in this press release include, among others, statements regarding our interpretation of preclinical studies, development plans for our product candidate, including the development of dose forms, the clinical trial testing schedule and milestones, the ability to complete enrollment in our clinical trials, interpretation of scientific basis for ganaxolone use, timing for availability and release of data, the safety, potential efficacy and therapeutic potential of our product candidate and our expectation regarding the sufficiency of our working capital. Forward-looking statements in this release involve substantial risks and uncertainties that could cause our clinical development programs, future results, performance or achievements to differ significantly from those expressed or implied by the forward-looking statements. Such risks and uncertainties include, among others, the uncertainties inherent in the conduct of future clinical trials, the timing of the clinical trials, enrollment in clinical trials, availability of data from ongoing clinical trials, expectations for regulatory approvals, the attainment of clinical trial results that will be supportive of regulatory approvals, and other matters, including the development of formulations of ganaxolone, and the availability or potential availability of alternative products or treatments for conditions targeted by the company that could affect the availability or commercial potential of our drug candidates. Marinus undertakes no obligation to update or revise any forward-looking statements. For a further description of the risks and uncertainties that could cause actual results to differ from those expressed in these forward-looking statements, as well as risks relating to the business of the Company in general, see Marinus' filings with the U.S. Securities and Exchange Commission. You may access these documents for free by visiting EDGAR on the SEC web site at www.sec.gov.

Ganaxolone: a positive allosteric GABAA modulator with a well-defined MOA designed to treat patients suffering from epilepsy and neuropsychiatric disorders

Clinical development includes late stage orphan diseases and large market opportunities with few or no treatment options

Multiple dose formulations IV and oral – to meet the ganaxolone needs of adult and pediatric patients in acute and chronic care settings

Extensive safety record in more than 1,600 subjects both pediatric and adult, at therapeutically relevant dose levels for up to two years

GNX is a synthetic analog of GNX is designed to modulate both synaptic and extrasynaptic GABAA allopregnanolone receptors to calm over-excited neurons

GANAXOLONE PHASE 1 PHASE 2 PHASE 3 MILESTONES

Seizure Disorders

Status Epilepticus (SE) Initiate mid-2020

CDKL5 Deficiency Disorder (CDD) Data Q3 2020

PCDH19-Related Epilepsy (PCDH19) Data 2021

Tuberous Sclerosis Complex (TSC) Initiate 1H 2020

Depressive Disorders*

Severe Postpartum Depression (PPD)

Treatment Resistant Depression (TRD)

*Pending regulatory interactions and funding

Cause Brain trauma, tumor, stroke, infection, among others

Seizure lasting >5 minutes or several seizures within 5 minutes with no recovery between Symptoms seizures

~45,000 patients fail 1st line treatment (US)1, 2 Prevalence Orphan Drug Designation

Few treatments available Treatments Benzodiazepines (1st line) Phenytoin/Fosphenytoin, valproic acid, levetiracetam, lacosamide (2nd line)

Mechanistic Synaptic receptors internalized and unavailable with prolonged seizures - need to Rationale modulate extrasynaptic receptors

1 Bleck et al. Epilepsia, 54(Suppl. 6):89–92, 2013 doi: 10.1111/epi.12288 2 LexisNexis PxDx Medical Claims data 2015 7 ©2020 Marinus Pharmaceuticals. All Rights Reserved. Goals of a New Therapy for the Treatment of SE

Established Status Refractory Status Super Refractory Status Epilepticus (ESE) Epilepticus (RSE) Epilepticus (SRSE)

Medically induced Coma

1st line 2nd line 3rd line

Benzodiazepine IV AED’s IV GNX IV Anesthetics Administered

Goals of a new treatment for SE Prevent patient progression towards escalation of treatment (IV anesthetics) Rapid cessation of status Maintenance of seizure control over study period

Diagnosis of convulsive or non-convulsive SE RSE Patients Failed at least one 2nd line IV AED but had not progressed to 3rd line IV anesthetics

Treatment Period Post-treatment Follow-up Screening Loading Dose Maintenance Taper 24 hr Weeks 2, 3, 4 Bolus plus 2-4 day 18 hour continuous infusion infusion taper

Cohort Dose of GNX/day N Low 500mg/day 5 Medium 650mg/day 4 Target 713mg/day 8

Endpoints • Primary: number of patients who do not require escalation of treatment with IV anesthetic within the first 24 hours after ganaxolone initiation • Secondary: additional efficacy, safety and tolerability

Dose of Last IV AED Dosing History of Failed Antiseizure Patient Etiology Type of SE Administered Prior to GNX 17 patients enrolled Cohort Epilepsy Medications Prior to GNX* (Recommended Dose) • 8 males, 9 females 1 Low Vascular No NCSE LAC, LEV 200mg (200-600mg) • Mean age: 57 years old (range: 23- 88) 2 Low Unknown Yes NCSE fPHT, LEV 1,000mg (1000-3000mg) 3 Low Vascular No NCSE LOR, LAC, LEV 600mg (200-600mg)

Types of SE 4 Low Vascular No NCSE LOR, LAC, LEV 600mg (200-600mg) • 5 (29%) CSE, 11 (65%) NCSE, 1 5 Low Tumor No CSE LOR, LAC, LEV 2,000mg (1000-3000mg) (6%) CSE→NCSE 6 Medium Vascular No NCSE LOR, LAC, LEV 600mg (200-600mg) Drug Overdose / History of Epilepsy 7 Medium Yes CSE LOR, LEV 1,000mg (1000-3000mg) Withdrawal • 7 (41%) yes, 10 (59%) no 8 Medium Unknown Yes CSE → NCSE LOR, LAC, LEV 1,000mg (1000-3000mg) Mean # of failed IV AEDs 9 Medium Tumor Yes NCSE LAC, LEV, PHT 200mg (100mg) (including benzodiazepines) 10 Target Vascular Yes CSE LOR, LAC, VPA 400mg (200-600mg) Drug Overdose / • 2.9 (range: 2-5) 11 Target No CSE LOR, LAC, LEV 400mg (200-600mg) Withdrawl Mean # of failed 2nd-line IV AEDs 12 Target Tumor Yes NCSE LOR, LEV, VPA 700mg (1000-3000mg) • 2.1 (range: 1-4), all failed LEV or 13 Target Autoimmune No NCSE LOR, LEV 1,000mg (1000-3000mg) LAC 14 Target Vascular No NCSE LOR, LAC, LEV, PHT 200mg (200-600mg)

• All prior AED’s were administered 15 Target Vascular Yes CSE LOR, LEV 1,000mg (1000-3000mg) within recommended dosing 16 Target Tumor No NCSE LOR, LAC, LEV 400mg (200-600mg) guidelines 17 Target Autoimmune No NCSE LOR, fPHT, LAC, LEV, VPA 200mg (200-600mg)

Data presented at AES 2019 *Bolded, underlined IV AED’s were the last ones administered prior to GNX 10 NCSE: Non-convulsive status epilepticus, CSE: Convulsive status epilepticus, LAC: Lacosamide, ©2020 Marinus Pharmaceuticals. All Rights Reserved. LEV: Levetiracetam, LOR: Lorazepam, PHT: Phenytoin, fPHT: Fosphenytoin, VPA: Valproic Acid Preliminary Data – Prior Two Failed Antiseizure Drugs (ASDs) Two previous ASD’s administered before GNX initiation )

d 24 e

t 1st Prior ASD a i t i

) 2nd Prior ASD n i

s s a r w

h o 18 PHT D ( t S

r n n e o o h i i w r

t LOR LOR s P e a 12 t

t LEV a i

c PHT e i LEV d n n

I LAC i LAC m

r LEV LEV i LEV a LEV X b T LEV

h 6 LOR N c LOR LOR LOR a

e LOR

G LAC f LAC LAC MDZ LAC VPA PHT LEV o LAC LAC fPHT p LOR LOR LAC LAC o t ( 0 LEV LOR LOR LEV LOR 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Low Medium Target Patient # Immediate previous ASD All prior ASD’s were administered at administered 4 hours (mean) therapeutic levels and within prior to GNX recommended dosing guidelines

11 LAC: Lacosamide, LEV: Levetiracetam, LOR: Lorazepam, PHT: Phenytoin, ©2020 Marinus Pharmaceuticals. All Rights Reserved. fPHT: Fosphenytoin, VPA: Valproic Acid, MDZ: Midazolam Ganaxolone Provides Rapid Onset and Durability of Efficacy

No escalation to additional IV AEDs or No escalation to IV IV anesthetics for anesthetics within 24 status relapse at any hrs from infusion Status-free through 24 time through 24 hrs No SE Relapse at initiation hrs from infusion after GNX anytime during the 4-wk SE Cessation Occurred Cohort (Primary Endpoint) initiation discontinuation follow up period Rapidly in All Dose Groups (median = 5 minutes) Target 100% 88% 100% 100% (713 mg/day) 300 (8 of 8) (7 of 8) (8 of 8) (6 of 6) (1ET, 1 died) (n=8) 200 30 Medium 100% 100% 75% 67% (650 mg/day) 20 (4 of 4) (4 of 4) (3 of 4) (2 of 3) (1 ET) (n=4) 10 Low 100% 100% 60% 50% (500 mg/day) (5 of 5) (5 of 5) (3 of 5) (1 of 2) (1 died) 0 (n=5)

ET: Early termination Target dose: 1 patient had status relapse @ D1 (resolved during the GNX infusion w/out tx escalation), 1 patient discontinued early (severe sedation on D1) & 1 patient died 9 days post-GNX discontinuation (bowel perforation, not considered drug related). Medium dose: 1 patient experienced status relapse @ D2 (during taper) & 1 patient discontinued on D3 (severe sedation). Low dose: 2 patients escalated to 3rd line tx for seizure relapse at D3 (1 died 16 days post-GNX discontinuation due to life support withdrawal & 1 patient died 22 days post-GNX discontinuation due to sepsis. None of the deaths were considered to be drug related. 12 Data presented at AES 2019 ©2020 Marinus Pharmaceuticals. All Rights Reserved. PK/PD Relationship and Rationale for Target Dose

Modeled PK Curves Seizure Burden Reduction Occurred Rapidly for All Dose Groups in All Dose Groups

Low Medium Target 500mg/day 650mg/day 713mg/day

) 0 d o i r G e n GNX E p e -20 e

E initiation d Dashed lines are PK models m r i n t

i u l

-40 a t B e

o t g

e /

r n e -60 u a m i z t h

i e e C r

u -80 S z i % e s ( -100

-1 0 1 2 4 6 8 10 12 14 Time (hrs) Target Dose Achieves ≥ 500 Only Target Dose Provided Sustained Reduction (>80%) ng/mL for ~8 hours Throughout Entire Analysis Window Data presented at AES 2019 13 ©2020 Marinus Pharmaceuticals. All Rights Reserved. IV Ganaxolone Safety Summary

10 SAEs in 6 patients (also included in AEs) 50 AEs in 16 patients

2 related in 2 patients 13 related in 7 patients

• 2 severe sedation • 6 mild (2 hypotension, 2 somnolence, 1 urinary retention, 1 hypercarbia) • 5 moderate (4 somnolence; 1 hypercarbia) • 2 severe (2 sedation)

8 non-related in 4 patients 37 not-related in 12 patients

• 1 Death due to withdrawal of life support • 20 mild – 1 Respiratory depression • 8 moderate (2 pain; 2 pneumonia, 2 dysphagia, • 1 Bowel perforation (fatal) • 1 delirium, 1 hypertension) • 1 Sepsis (fatal) • 9 severe (respiratory depression, death due to withdrawal of • 1 Fall support, sepsis, embolic stoke, perforated bowel, fall, loss of – 1 Loss of consciousness consciousness, multiple fractures, pneumothorax) – 1 Pneumothorax – 1 Multiple fracture Intubation • 9 patients were not intubated upon enrollment. Of these, 6 remained intubation-free during the entire ganaxolone treatment period

Study Design Randomized, placebo-controlled (adjunctive to standard-of-care) clinical trial

Refractory SE patients with similar baseline characteristics (etiology, # of failed IV Target Patient Population AED’s, etc.) to the Phase 2 study

Dosing: Target dose range but extends GNX exposure ≥ 500 ng/mL from 8 to 12 hours

Primary and Key Focuses on clinically-meaningful effects related to rapid onset of action, durability Secondary Endpoints of effect, and prevention of treatment escalation

Target key areas of safety (e.g., % intubated on GNX, time to extubation, etc.) and Other Endpoints healthcare utilization metrics (e.g., length of stay, # of days in the ICU, etc.)

Targeting End of Phase 2 meeting with FDA in Q1-2020

*Proposed Phase 3 clinical trial elements contingent on future regulatory interactions* 15 ©2020 Marinus Pharmaceuticals. All Rights Reserved. CDKL5 Deficiency Disorder (CDD) – Rare, Serious Epileptic Condition

Mutation of the cyclin-dependent kinase-like 5 (CDKL5) gene, located on the X Cause chromosome

Early-onset, treatment refractory seizures, & severe neuro-developmental delay Symptoms Most can’t walk, talk or care for themselves Suffer from scoliosis, visual impairment, gastrointestinal difficulties & sleeping disorders

~12,500 children US and EU5, predominantly affects females Prevalence Genetic testing available Orphan Drug designation

Treatments No approved treatments

Mechanistic Potential GABA dysfunction Rationale A

Treatment Open-Label Phase Baseline 26 weeks 52 weeks 12 weeks Ganaxolone 600 mg 3x/day maximum 600 mg 3x/day maximum

Trial Details Baseline Characteristics Endpoints 2 sites in US; 1 site in Italy Mean number seizures – 206 Primary: % change in seizure (range 34 to 669) frequency per 28 days relative to baseline 6 females, 1 male – ages 2-16 Median number seizure-free days - 4 Secondary: % increase in seizure (range 0 to 9) free days from baseline, safety and tolerability, CGI Confirmed CDKL5 mutation, stable background treatment, >4 seizures per 28-day period in baseline

• Patients experienced a median ‘minimally improved’ rating on CGI-I scale (Clinician rated) correlated 44.4% with % change in seizure frequency Reduction in seizure • GNX was generally safe and well-tolerated (no SAEs). No reports of somnolence or dizziness. frequency − 2 of the 7 patients discontinued prior to completing treatment due to lack of efficacy

Correlation between change in seizure frequency and CGI-I scale

1: very much improved 2: much improved 3: minimally improved 4: no change 5: minimally worse 6: much worse 7: very much worse

Very much improved no change Very much worse

• 4 of 7 patients entered the extension period

• Ganaxolone demonstrated preliminary evidence of sustained, long-term (out to 18 months) efficacy in a small cohort of CDD patients

• Efficacy of existing AEDs and the ketogenic diet in patients with the CDKL5 mutation is low and the durability of effect is short1

Patients Entering OLE 54% 66% Median change in seizure Median change frequency frequency at 6 months improved to 66% in extension period (as of 12/31)

Clinical data presented at AES 2018 1Müller, A. et al., European Journal of Paediatric Neurology, Volume 20 (2016), 147-151. 19 ©2020 Marinus Pharmaceuticals. All Rights Reserved. Global Phase 3 Pivotal Study Design

Ganaxolone (600 mg, 3x day maximum) R Ganaxolone n=100 1:1 (600 mg, 3x day maximum)

Placebo

Historical Control Baseline Titration Maintenance Titration Open-Label Phase 8 weeks 6 weeks 4 weeks 13 weeks 4 weeks

Double-Blind Phase Open-Label Phase

Trial Details Endpoints ~43 global sites; US, Europe & others Primary Endpoint: % change in seizure frequency

Ages 2-21, 16 major motor drop seizures/month; up to 4 Non-seizure secondary outcome measures: concomitant AEDs Behavioral/neuropsychiatric changes correlated with domains of attention & sleep

Cause Inherited mutation of protocadherin 19 (PCDH19) gene. Located on X chromosome

Early-onset seizures, cognitive and sensory impairment, & psychiatric and behavioral Symptoms disorders. Seizures last from days to weeks; sensitive to fever Affects ~10-12K children US and EU5 Predominantly females Prevalence Genetic testing becoming more readily available Orphan Drug designation

Treatments No approved treatments. No previous clinical trials.

Mechanistic Associated with low levels of allopregnanolone1 and potential GABA dysfunction Rationale A

• 25% decrease in median seizure frequency reported in Phase 2 (n=11) • Stratification of patients by baseline plasma Allo levels identifies a subpopulation with improved efficacy on GNX • When comparing seizure frequency at 6 months to baseline: − The biomarker+ group (low Allo) significantly improved − The biomarker- group (high Allo) deterioration was not statistically significant

Median % change in P-value Biomarker N= seizure rates (Wilcoxon)

Biomarker + 7 -50% P=0.02 (Allo-S <2,500 pg mL-1)

Biomarker – 4 84% P=0.63 (Allo-S >2,500 pg mL-1)

Clinical data presented at AES 2018 22 ©2020 Marinus Pharmaceuticals. All Rights Reserved. Biomarker Stratified Phase 3 Pivotal Study in PCDH19 Primary efficacy analyses will be conducted using this cohort

Low Allo-S Placebo Biomarker + R n = 50 1:1 Ganaxolone Up to 600 mg 3x/day PCDH19 (all-patients) Ganaxolone n~ 70 Up to 600 mg 3x/day High Allo-S Placebo Biomarker - R n~ 20 1:1 Ganaxolone Up to 600 mg 3x/day

Open-Label Phase (52 weeks) Pre-baseline Screening Baseline 12 weeks Titration 4 wks Maintenance 13 weeks Titration 4 wks * not drawn to scale

Screening Visit Trial Details Endpoint ~35 global sites; US, Europe & others Primary Endpoint: % change in seizure frequency Ages 1-17, 8 or more seizures in 8 weeks; failed 2 or more AEDs Data expected 2021

TSC

Cause Defect or mutation of TSC1 and/or TSC2 genes

Symptoms Benign tumors, seizures, cognitive impairment, behavioral problems, skin abnormalities

Prevalence ~25K patients with refractory epilepsy

Treatments Limited approved treatments

Mechanistic GABA receptor active steroids are altered1 Rationale A

PART A PART B

Open-Label Extension (OLE) Base GNX Titration GNX Maintenance (24 Weeks) (4 Weeks) (4 Weeks) (8 Weeks) * Available to patients that respond to GNX as defined per protocol

Baseline Period Treatment Period OLE Period

Screening Baseline 2-week taper upon GNX discontinuation Visit Treatment (if not continuing to Part B) Visit

• n = 30

• 4 - 5 U.S. sites

• Primary efficacy endpoint: % change in primary seizure frequency Primary seizure types: focal motor seizures without impairment of consciousness or awareness, focal seizures with impairment of consciousness or awareness, focal seizures evolving to bilateral generalized convulsive seizures, and generalized seizures with a motor component that are countable

Status Epilepticus, Pediatric Seizure Disorders, Depression

Attractive orphan indications & large market indications

Significant unmet medical need – no or few treatment options

Targeted patient populations – understood dysfunctional GABA system

IV to convenient oral – continuity of treatment from hospital to outpatient transition

New MOA, Differentiated, Convenient Dosing, Targeted CNS Therapy