Purpose
– The purpose of this paper is to investigate the patterns and determinants of excess labour turnover (churning) in the Luxembourg labour market using a rich employer‐employee matched data set.

Design/methodology/approach
– The paper formulates a model to explain churning rates using three sets of explanatory variables: various worker characteristics, establishment characteristics, and two‐digit sector‐specific characteristics. The data used are from the Luxembourg social security system for the period 1992‐2003.

Findings
– The findings show that there are high churning levels in Luxembourg, that their determinants vary significantly across sectors, and that much of this variation can be explained by worker‐ and establishment‐specific features.

Research limitations/implications
– A major question, still undecided in the research literature, is whether churning is simply the result of random job‐worker mismatches.

Originality/value
– There are relatively few prior studies that have examined which employee and employer characteristics are associated with excess worker turnover, and the paper is the first to analyze the phenomenon of churning in the Luxembourg labour market. Luxembourg has recently experienced impressive employment growth (about 3.5 percent annually) since the beginning of the 1990s; and, Luxembourg being a small economy, it was feasible to analyze the entire population of workers and firms.