Abstract This paper presents a classification of online quality control (QC) methods from the view of process control. The QC can be applied on the control loops from each of its three sides: the input (manipulative variable) side, the output (controlled variable) side and from the disturbance side. It was found that online QC can be direct or indirect, depending on the measures taken for quality. This classification can lead to interesting and new options for the control variables that otherwise would have been obscure. Once the proper control variable is selected (in terms of adequate representation of quality) it can be used for control systems analysis and design. Process application is presented for an industrial Steam Gas Reformer. The input is the fuel gas quality for which various options were presented. A correlation was obtained to relate heat input to simple measurements. The output hydrogen quality control options were discussed. Coil outlet temperature is adequate for a crude estimate of conversion, provided that S/C ratio is controlled. S/C ratio correlation was obtained to enable its estimation and control from simple measurements. A precise quality control of hydrogen can be achieved provided that COT is also controlled to protect reformer catalyst. An improved strategy can be implemented where both COT and conversion are controlled in a multivariable sense. This strategy is economically attractive, since it allows continuous manipulation of S/C ratio to the minimum required for COT control. Savings in fuel gas can be achieved accordingly. The feasibility of multivariable control was established via interaction analysis.