Due to the complexity of IMRT dosimetry, dose delivery evaluation is generally done using a treatment plan in which the optimized fluence distribution has been transferred to a test phantom for accessibility and simplicity of measurement. The actual patient doses may be reconstructed in vivo through the use of electronic portal imaging devices or films, but the assessment of absolute dose from these measurements is time-consuming and complicated. In our clinic we have instituted the use of routine diode dosimetry for IMRT patients following the same procedure used for standard radiation therapy patients in which each new treatment field is checked at the start of treatment. For standard cases the dose at dmax is calculated as part of the monitor unit calculation. For the IMRT cases, the dose contribution to the dmax depth for each field is taken from the treatment plan. We found that about 90% of the diode measurements agreed to within +/- 10% of the planned doses (45/51 fields) and 63% (32/51 fields) achieved +/- 5% agreement. By using this direct in vivo method to verify the clinical doses delivered, we have been able to make a uniform startup procedure for all patients while simplifying our IMRT QA process.