This article investigates the efficiency of the matching process by panel stochastic frontier estimation of the matching function in Croatia. The empirical analysis is conducted on a regional level using regional office-level monthly data obtained from the Croatian Employment Service for 2000–2011. The results suggest that the efficiency of the matching process is rising over time, although with significant regional variations. In order to explore these variations, structural characteristics of the labour market together with some policy variables are included in the second-stage estimation. Various structural variables have different impact on the matching efficiency, while policy variables are mostly positively correlated with it. For instance, both active labour market programs and the number of high-skilled employees in regional employment offices positively affect matching efficiency. Additionally, when regional income per capita is included in the model, it shows positive impact on the matching efficiency, indicating that demand fluctuations predominantly affect the matching process. Finally, in order to get consistent estimates, panel stochastic frontier model transformation is applied. The obtained results show that there is no major difference in estimated mean technical efficiency coefficients in comparison to the original estimation.