Physical inactivity is a major contributor to non-communicable disease and people of low socioeconomic status (SES) are more likely to be insufficiently active. Physical activity mass media campaigns aim to increase physical activity participation, but little is known about their impact on low SES groups. We reviewed the published literature from 1990 to June 2016 to identify reports of physical activity mass media campaigns. We documented evaluation/study design, target population, campaign outcomes assessed, SES measures used and analysed the results of the SES comparisons. A total of 23 papers were reviewed, reporting on 17 physical activity campaigns and 12 campaigns compared SES differences for 85 outcomes: 45 comparisons showed no difference between lowest and highest SES groups, 20 showed a better outcome for the lowest SES group and 20 showed a worse outcome. Some campaigns found inconsistent results, but seven found only equal and/or better results for low SES groups. Post-campaign physical activity behaviour most commonly showed no SES differences, but no other patterns were seen. Our review found that physical activity mass media campaigns have mostly equitable or better impacts for low SES groups, but to reduce inequalities these campaigns need to be maximally effective for low SES populations.