We introduce PoPreRo, the first dataset for Popularity Prediction of Romanian posts collected from Reddit. The PoPreRo dataset includes a varied compilation of post samples from five distinct subreddits of Romania, totaling 28,107 data samples. Along with our novel dataset, we introduce a set of competitive models to be used as baselines for future research. Interestingly, the top-scoring model achieves an accuracy of \(61.35\%\) and a macro \(F_1\) score of \(60.60\%\) on the test set, indicating that the popularity prediction task on PoPreRo is very challenging. Further investigations based on few-shot prompting the Falcon-7B Large Language Model also point in the same direction. We thus believe that PoPreRo is a valuable resource that can be used to evaluate models on predicting the popularity of social media posts in Romanian. We release our dataset at https://github.com/ana-rogoz/PoPreRo .