The research of my group is primarily concerned with the study of star forming galaxies. On the observational side, studies include the relationship between star formation and the ISM, including molecular, and the neutral and ionized atomic components. A particular interest is the outflowing winds of starburst galaxies, notably M82, where a number of detailed observations have been made. A current project on M82 is the preparation of a map in the high excitation line CO J=6-5 with the JCMT to investigate the distribution of the hottest and most dense gas, particularly in the vicinity of the superbubble in the south-west part of the disk. Another recently completed study is the investigation of the conditions in the gas and dust of the Antennae interacting pair. A similar study of the "Taffy Galaxies" is currently underway. We have conducted a survey of 60 of the 104 galaxies in the "SLUGS" galaxy sample at CO J=1-0 (Nobeyama) and CO J=3-2 (JCMT), both at the same resolution. Preliminary conclusions are (1) that the 850 micron continuum emission measured by SCUBA is significantly contaminated by CO J=3-2 line emission; (2) that based on modelling by LVG methods, the CO is over-luminous by an order of magnitude compared to Galactic GMC's; and (3) that the warm gas/dust ratios in the nuclear regions are about 100 with an uncertainty of a factor of two. Another area I am supporting is the study of the molecular and atomic gas at the interface between the disk and haloes of star-forming galaxies. In a study led by Siow-Wang Lee, we have just completed an investigation of a CO and HI emitting super-shell in the edge-on galaxy NGC 5775. On the theoretical side, graduate student Lihong Yao and I are developing a model for a starburst galaxy. The aim is to first model the interaction between newly formed star clusters and the expanding shells of molecular gas propelled by winds and supernovae. Upon this framework, we will incorporate PDR models which will predict time dependence of the molecular line emission from a variety of species and excitations. The long term goal is to determine whether different phases of a starburst can be identified by the molecular and atomic line properties in a manner analagous to the use of IR emission to model starbursts. A secondary aim is to model the relationship between molecular/atomic line emission and molecular hydrogen content to better determine the good and bad tracers of molecular hydrogen. In particular, we hope to understand more clearly the reason why the widely adopted "X-factor" used to trace molecular hydrogen by 12CO J=1-0 emission in the Milky Way overestimates the molecular gas in starburst conditions.