Invading plants remain undetected in a lag phase while they explore suitable climates.

Successful alien species may experience a period of quiescence, known as the lag phase, before becoming invasive and widespread. The existence of lags introduces severe uncertainty in risk analyses of aliens as the present state of species is a poor predictor of future distributions, invasion success and impact. Predicting a species' ability to invade and pose negative impacts requires a quantitative understanding of the commonality and magnitude of lags, environmental factors and mechanisms likely to terminate lag. Using herbarium and climate data, we analysed over 5,700 time series (species × regions) in 3,505 naturalized plant species from nine regions in temperate and tropical climates to quantify lags and test whether there have been shifts in the species' climatic space during the transition from the lag phase to the expansion phase. Lags were identified in 35% of the assessed invasion events. We detected phylogenetic signals for lag phases in temperate climate regions and that annual self-fertilizing species were less likely to experience lags. Where lags existed, they had an average length of 40 years and a maximum of 320 years. Lengthy lags (>100 years) were more likely to occur in perennial plants and less frequent in self-pollinating species. For 98% of the species with a lag phase, the climate spaces sampled during the lag period differed from those in the expansion phase based on the assessment of centroid shifts or degree of climate space overlap. Our results highlight the importance of functional traits for the onset of the expansion phase and suggest that climate discovery may play a role in terminating the lag phase. However, other possibilities, such as sampling issues and climate niche shifts, cannot be ruled out.