The detailed investigation of the broadband flux variability in the blazar 3C 273 allowed us to probe the location and size of emission regions and their physical conditions. We conducted correlation studies of the flaring activity in 3C 273, which was observed for the period between 2008 and 2012. The observed broadband variations were investigated using the structure function and the discrete correlation function methods. Starting from the commonly used power spectral density (PSD) analysis at X-ray frequencies, we extended our investigation to characterise the nature of variability at radio, optical, and γ -ray frequencies. The PSD analysis showed that the optical and infrared light-curve slopes are consistent with the slope of white-noise processes, while the PSD slopes at radio, X-ray, and γ -ray energies are consistent with red-noise processes. We found that the estimated fractional variability amplitudes strongly depend on the observed frequency. The flux variations at γ -ray and mm-radio bands are found to be significantly correlated. Using the estimated time lag of (110 ± 27) days between γ -ray and radio light-curves, where γ -ray variations lead the radio bands, we constrained the location of the γ -ray emission region at a de-projected distance of 1.2 ± 0.9 pc from the jet apex. Flux variations at X-ray bands were found to have a significant correlation with variations at both radio and γ -ray energies. The correlation between X-ray and γ -ray light curves indicates two possible time lags, which suggests that two components are responsible for the X-ray emission. A negative time lag of −(50 ± 20) days, where the X-rays are leading the emission, suggests that X-rays are emitted closer to the jet apex from a compact region (0.02–0.05 pc in size), most likely from the corona at a distance of (0.5 ± 0.4) pc from the jet apex. A positive time lag of (110 ± 20) days ( γ -rays are leading the emission) suggests a jet-base origin of the other X-ray component at ~4 to 5 pc from the jet apex. The flux variations at radio frequencies were found to be well correlated with each other such that the variations at higher frequencies are leading the lower frequencies, which is expected from the standard shock-in-jet model.