The ACE-2 HILLCLOUD experiment was carried out on the island of Tenerife in June-July 1997 to investigate the interaction of the boundary layer aerosol with a hill cap cloud forming over a ridge to the north-east of the island. The cloud was used as a natural flow through reactor to investigate the dependence of the cloud microphysics and chemistry on the characteristics of the aerosols and trace gases entering cloud, and to simultaneously study the influence of the physical and chemical processes occurring within the cloud on the size distribution, chemical and hygroscopic properties of the aerosol exiting cloud. 5 major ground base sites were used, measuring trace gases and aerosols upwind and downwind of the cloud, and cloud microphysics and chemistry and interstitial aerosol and gases within the cloud on the hill. 8 intensive measurement periods or runs were undertaken during cloud events, (nocturnally for seven of the eight runs) and were carried out in a wide range of airmass conditions from clean maritime to polluted continental. Polluted air was characterised by higher than average concentrations of ozone (> 50 ppbv), fine and accumulation mode aerosols (>3000 and >1500 cm-3, respectively) and higher aerosol mass loadings. Cloud droplet number concentrations N, increased from 50 cm-3 in background maritime air to >2500 cm-3 in aged polluted continental air, a concentration much higher than had previously been detected. Surprisingly, N was seen to vary almost linearly with aerosol number across this range. The droplet aerosol analyser (DAA) measured higher droplet numbers than the corrected forward scattering spectrometer probe (FSSP) in the most polluted air, but at other times there was good agreement (FSSP = 0.95 DAA with an r2 = 0.89 for N < 1200 cm-3). Background ammonia gas concentrations were around 0.3 ppbv even in air originating over the ocean, another unexpected but important result for the region. NO2 was present in background concentrations of typically 15 pptv to 100 pptv and NO3. (the nitrate radical) was observed at night throughout. Calculations suggest NO3. losses were mainly by reaction with DMS to produce nitric acid. Low concentrations of SO2 (apprx30 pptv), HNO3 and HCl were always present. HNO3 concentrations were higher in polluted episodes and calculations implied that these exceeded those which could be accounted for by NO2 oxidation. It is presumed that nitric and hydrochloric acids were present as a result of outgassing from aerosol, the HNO3 from nitrate rich aerosol transported into the region from upwind of Tenerife, and HCl from sea salt aerosol newly formed at the sea surface. The oxidants hydrogen peroxide and ozone were abundant (i.e., were well in excess over SO2 throughout the experiment). Occasions of significant aerosol growth following cloud processing were observed, particularly in cleaner cases. Observations and modelling suggested this was due mainly to the take up of nitric acid, hydrochloric acid and ammonia by the smallest activated aerosol particles. On a few occasions a small contribution was made by the in-cloud oxidation of S(IV). The implications of these results from HILLCLOUD for the climatologically more important stratocumulus Marine Boundary Layer (MBL) clouds are considered.