iFit is a new intensity-based retrieval algorithm for direct fitting of measured UV spectra, specifically developed for use in volcanology. It has been designed with a focus on minimising processing of the measured spectra prior to analysis. Here, we report a detailed presentation of the iFit algorithm, and test it in 4 case studies, examining clear sky spectra, SO<sub>2</sub> calibration cell experiments and volcanic SO<sub>2</sub> and BrO retrievals from traverse measurements performed on Mt. Etna volcano, Italy. We show that the major source of fit error in the intensity fitting come from air mass factor independent solar spectrum errors, which are, however, easily characterised and removed by dividing the measured spectrum by a pre-calculated solar spectrum residual. Furthermore, we have quantified the magnitude of the flat spectrum in two spectrometers, and shown that this spectrum is strongly spectrometer dependent but temperature independent, opening the possibility for robust analysis and and BrO without the need for temperature stabilisation of quantification of both SO<sub>2</sub> the spectrometer. We find that iFit can be robustly and easily applied to traverse measurements of volcanic plumes, producing bias-free profiles of SO<sub>2</sub> , and high quality SO<sub>2</sub> /BrO ratios without the need for clear sky background spectra. Fit residuals are typically pure instrumental noise when the residual solar spectrum is removed. We believe that the iFit approach, which avoids the need for a clear sky spectrum and which achieves noise-limited fits, is ideally suited to the automated analysis of spectra produced by networks of scanning UV spectrometers around volcanoes.