CBFC Watch

Trends and analysis

To give you a sense of what's in the data, we've highlighted a few key trends and relationships below. These visualizations are designed to illustrate the types of questions this dataset can help answer and to serve as a starting point for analysis.

This page may contain explicit terms and offensive language from film censorship records.

CBFC modification rates show consistent patterns for violence, volatile trends for cultural topics

Film censorship varies most for religious and political content

Profanity

-50%+0%+50%+100% ; ; ; ; ; ; ; ; ; ; ; -50%+0%+50%+100%

Substance Use

; ; ; ; ; ; ; ; ; ↑ More than usual ↓ Less than usual

Violence

; ; ; ; ; ; ; ; ; ;

Sexual Content

-50%+0%+50%+100% ; ; -50%+0%+50%+100%

Religious Content

; ; ; ; ; ; ; ; ; ; ; ; ; ;

Political Content

; ; ; ; ; ; ; ; ; ; ;

Analysis of Central Board of Film Certification modification records, showing 3-month moving average of deviations from baseline censorship rates by content type. Baseline represents the overall average modification rate for each category from 2017-2025.

Cuttack and Delhi offices modify more content in films than other regions

The average amount of content modified varies by certifying office location.

Average minutes of content modified per movie0 sec5 m10 m HyderabadBangaloreKolkataChennaiThiruvananthpuramMumbaiDelhiCuttack HyderabadBangaloreKolkataChennaiThiruvananthpuramMumbaiDelhiCuttack 2.0 min2.2 min2.5 min3.1 min3.3 min3.8 min4.1 min9.2 min

Hindi leads film production, followed by Tamil and Telugu

Distribution of films by language in on E-CinePramaan (2017-2025).

Number of films02K4K GujaratiBengaliMarathiBhojpuriMalayalamEnglishKannadaTeluguTamilHindi GujaratiBengaliMarathiBhojpuriMalayalamEnglishKannadaTeluguTamilHindi 3664987371,0311,3381,6871,9752,5582,6714,178

Profanity and violence are the most commonly modified content types

Total number of modifications by content category across all films.

Number of modifications010K20K Identity ReferenceReligiousPoliticalSexual ExplicitSexual SuggestiveSubstanceViolenceProfanity Identity ReferenceReligiousPoliticalSexual ExplicitSexual SuggestiveSubstanceViolenceProfanity 1,5002,5892,6212,97511,63613,25421,14221,789

Heavily modified movies by office

Top ten movies from each CBFC office by total duration of modifications. Select an office to see the films with the most content removed.

Heavily modified movies

Films with the most content removed by CBFC offices

Office
Film
Duration
MumbaiMASS29m 52s
MumbaiDAROGA BABUNI29m 50s
MumbaiJIYAB SHAAN SE29m 46s
MumbaiDON ( DON SEENU)29m 44s
MumbaiNAKSHATRAM29m 43s
MumbaiPALAHATI BRAHMA NAIDU - SHEZADA29m 43s
MumbaiSAPTAMA INDRIYA29m 38s
MumbaiHAI29m 37s
MumbaiAATAADISTHA29m 35s
MumbaiDARINGBAAZ KHILADI RETURNS (THOOTTAL POO MALARUM)29m 24s

Distinctive keywords by modification category

Terms with the highest Term Frequency-Inverse Document Frequency (TF-IDF) score for each category. TF-IDF is a measure used to find words that are uniquely important to a specific topic. A high score means a term is a distinctive signal for that category.

CONTENTMEASURES
TYPE
WORD/PHRASE
COUNT
TF-IDF
religioussuperstition1220.0445
violenceblood26890.0315
substancetobacco4620.0285
sexual_suggestivekissing scene3340.0223
politicalnational flag390.0196
sexual_suggestivecleavage5800.0185
sexual_suggestivesexual_suggestive2680.0179
sexual_suggestivekissing8940.0175
substanceliquor label1990.0173
sexual_suggestivelove making scene1800.0170
substanceakshay kumar1800.0157
politicalmodi640.0152
religiousblack magic570.0147
sexual_explicitnaked woman570.0145
substanceliquor labels1110.0145
sexual_explicitrape scene810.0145
religiousreligion550.0142
sexual_explicitnudity1880.0140
substancerahul dravid1050.0137
substancesmoking and drinking1030.0135
Showing 1-20 of 300 results

Use this data

We’re making our code and data available so you can dig into the numbers yourself. Our RMarkdown notebook has everything you need to replicate our charts and analysis, or to use as a jumping-off point for your own projects. And if you spot a mistake in our work, please file an issue on our GitHub.

Data analysis notebook