This paper reports new developments in in situ U-Pb zircon geochronology using 266 and 213 nm laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS). Standard spot ablation (spot diameters 40-80 Am) was employed, with no sampling strategies employed specifically to minimise elemental fractionation. Instead, He ablation gas and carefully replicated ablation conditions were employed to maintain constant ablation-related elemental fractionation of Pb and U between analyses. Combining these strategies with calibration on a new zircon standard (GJ-1) allows elemental fractionation and instrumental mass bias to be corrected efficiently, and accurate 206Pb/238U and 207Pb/235U ratios to be measured with short-term precision (2 r.s.d.) of 1.9% and 3.0%, respectively. Long-term precision (2 r.s.d.) of the technique (266 nm ablation), based on 355 analyses of the 91500 zircon (1065 Ma) standard over more than a year, was 3.8%, 4.0% and 1.4% for the 206Pb/238U, 207Pb/235U and 207Pb/206Pb ratios, respectively. Long-term precision (2 r.s.d.) for the 206Pb/238U, 207Pb/235U and 207Pb/206Pb ratios of the Mud Tank zircon (732 Ma) was 3.9%, 4.1% and 1.7%, respectively (359 analyses). Selective integration of time-resolved signals was used to minimise the effect of Pb loss and common Pb enrichments on the measured ages. The precision and accuracy of our data compare very favourably with those obtained using more involved procedures to correct or minimise ablation- and ICP-MS-induced biases. 213 nm laser ablation produced comparable precision to 266 nm ablation using generally smaller spot sizes (40-50 vs. 60-80 Am), and offered significant advantages in terms of ablation duration and stability, particularly for small zircons (b60 Am). For the 91500 zircon, but not the Mud Tank zircon, 213 nm ablation also produced significantly older and more accurate Pb/U ages. This suggests that shorter wavelength ablation may have reduced a matrix-dependent elemental fractionation difference between sample and standard. The accuracy and precision of the technique for young zircons are demonstrated by analysis of three zircon populations ranging in age from 417 to 7 Ma. In each case, the zircons have yielded concordant ages or common Pb discordia which give concordia intercept ages that are in agreement with independently determined ages for the same samples. Application of Tera-Wasserburg diagrams [Earth Planet. Sci. Lett. 14 (1972) 281] was found to be the most useful approach to handling common Pb contributions that were not removed by selective integration of signals.