The London office market is one of the most researched market in the world owing both to size and data availability. Prior work modelling office market rental adjustment was based on only two decades of data that were dominated by less than an incomplete real estate cycle (Hendershott et al, 1999; 2002; Wheaton et al., 1997),with vacancy peaking in 1991 but not reaching its trough by 1996. Adding another decade completes the earlier cycle in 2000 and adds another full cycle (peak in 2003, trough in 2006). We use these expanded data to test whether earlier estimates of market adjustment are consistent with the most recent decade of London data. We then test for two possible asymmetric responses in market adjustment. First, positive and negative shocks to employment and supply may have different impacts on the adjustment process, given non-negativity constraints (Grenadier, 1995, 1996). Second, adjustment could vary depending on whether the vacancy rate is high or low when the shock occurs (Englund, et al., 2006). The paper sets out a model of the rental adjustment process, details the dataset employed then models rents using an error correction mechanism approach, with long run demand for office space being a function of real rent and employment and with an equilibrium vacancy rate. New lease rents adjust to changes in the market drivers and to the gaps between actual and natural vacancy rate and actual and equilibrium real rent. We then use this framework to seek for asymmetries in the adjustment process.