1. Triplicates? Yes. Keep triplicates for all standards and QCs, and use the mean after rejecting obvious outliers (e.g., Grubbs or a simple “>2 SD” rule). Triplicates let you spot edge effects and pipetting drift.

2. Same plate or separate?

Best: run GSH (or total GSH) and GSSG on the same plate, same session, if the kit’s workflow allows (i.e., includes the GSH-masking step for GSSG). This minimizes plate-to-plate variation and temperature/timing differences.

If you must split plates: run them back-to-back the same day and bridge with shared controls (low/mid/high QC prepared once, aliquoted across both plates) + duplicate standard level(s) on both plates to check for shifts.

3. Automatic injector + 30-s pauses? Yes, that can absolutely hurt GSSG reproducibility. Many colorimetric/enzymatic kits are time-sensitive per well. A staggered 30-s delay means wells have different incubation times, which shows up worst at low signals (often the GSSG curve). See “Timing & dispensing” below.

Best-practice checklist (GSH/GSSG)

Derivatize/mask GSH completely before GSSG readout. Reagents like N-ethylmaleimide (NEM) or 2-vinylpyridine must be fresh and at the kit-specified excess; under-masking lets GSH bleed into the GSSG signal.

Prevent thiol redox artifacts: prepare standards/samples in cold, slightly acidic conditions with EDTA (e.g., 0.5–5% sulfosalicylic acid or 1–10 mM HCl + 1–5 mM EDTA, per kit guidance). Keep everything on ice, minimize light/air exposure, and avoid metals.

No reducing agents anywhere (no DTT, β-ME, TCEP in buffers, tips, or leftover labware).

Matrix matching: if you’re quantifying plasma, prepare a second standard curve spiked into GSH/GSSG-stripped plasma (or dialyzed matrix) to check for matrix effects.

Range: build standards that tightly bracket your expected sample values; don’t over-stretch the curve. If the low end is noisy, start the curve slightly higher and dilute samples to land in the reliable range.

Mixing: mix gently but thoroughly after every addition; avoid bubbles (they distort optical reads).

Plate reader: pre-warm, use the recommended wavelength(s), consistent pathlength correction if applicable, and read at the same interval after reagent addition for every well.

Plate strategy & layout

Include: triplicate standards, reagent blank (no analyte), matrix blank, and 3 QCs (low/mid/high) made independently from standards.

Avoid edge effects: fill outer wells with buffer or leave empty; put standards/QCs in interior wells.

Map: S (standards, triplicate), Q (QCs), B (blanks), then samples in columns by treatment group to reduce timing skew.

Timing & dispensing (important for your injector)

Staggering hurts. If your injector adds to wells sequentially with long pauses, wells have unequal incubation times.

Fixes:

Use a multichannel pipette or a plate dispenser that can dose a full column/row quickly, then work column-by-column with a column timer (e.g., read Column 1 exactly X min after addition, Column 2 exactly X + 20–30 s, etc., and calculate times accordingly).

Or run a kinetic read immediately after reagent addition and fit the initial rate (slope) within a fixed time window (e.g., 1–3 min). Initial rates are less sensitive to small time offsets than endpoint signals.

If you keep the injector, reduce the delay to the minimum safe for overflow, and standardize the delay across all wells; process in consistent blocks (e.g., 4 columns at a time) and read blocks on schedule.

Temperature control: keep the plate on a temperature-controlled deck or incubator; small drifts change enzyme rates noticeably.

Curve fitting & acceptance

Don’t chase R² alone. Also check:

Back-calculated accuracy of standards (acceptable ~85–115%, 80–120% at the extremes).

%CV of replicates (typically

Note: